[CTN] CTN seminar: Dr. Surya Ganguli (Stanford), Oct 29, 3.30pm in PAS 2464

[Matthijs van der Meer]

Dear all,

Please join us for next Tuesday's CTN seminar (Oct 29) by Dr. Surya
Ganguli, from Stanford. Title and abstract follow below.

Time and place are the usual, 3.30pm on Tuesday in PAS 2464.

If you would like to meet with Dr. Ganguli, and/or come to dinner
(on Tuesday evening), please let me know.

Hope to see you there,

- Matt


Dr. Surya Ganguli
Dept. of Applied Physics,
and, by courtesy,
Dept. of Neurobiology and
Dept. of Electrical Engineering

Stanford University

Title: A theory of neural dimensionality and dynamics

Abstract: In a wide variety of experimental paradigms, neuroscientists
tightly control behavior, record many trials, and obtain trial averaged
neuronal firing rate data from hundreds of neurons, in circuits
containing millions to billions of behaviorally relevant neurons. Such
datasets are often analyzed by dimensionality reduction methods that
allow us to visualize neuronal dynamics through their projections onto a
number of basis patterns. Strikingly, recordings from hundreds of
neurons can often be described using a much smaller number of dimensions
(basis patterns), and the resulting projections yield a remarkably
insightful dynamical portrait of neural circuit computation. Thus many
neuronal datasets are surprisingly simple, and we seem to be able to
extract reasonable collective neuronal dynamics despite overwhelming
levels of neuronal subsampling. This ubiquitous simplicity raises
several profound and timely conceptual questions. What is the origin of
this simplicity? What does it tell us about the complexity of brain
dynamics? Would neuronal datasets become more complex if we recorded
more neurons? How and when can we trust dynamical portraits obtained
from only hundreds of neurons in a circuit containing billions of
neurons? More generally, what, if anything, can we learn about a complex
dynamical system by measuring an infinitesimal fraction of its degrees
of freedom? We present a theory of neural dimensionality and dynamics
that answers all of these questions, and we further test this theory in
neural recordings from monkeys performing reaching movements.